Modeling methods for tidal flat digital terrain based on neural network

Author

Li, Qing ; Ding, Xianrong ; Zhu, Ang ; Cheng, Ligang ; Kang, Yanyan ; Ge, Xiaoping ; Zhang, Jing

Author_Institution

State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjng 210098, China

fYear

2010

fDate

4-6 Dec. 2010

Firstpage

6716

Lastpage

6719

Abstract

The purpose of this paper is to model tidal flat digital terrain. The study tidal flats are in the yellow sea radial sand ridges eastern China. Based on the regularity and variability characteristics of changeable tidal flats, combined with remote sensing and remote surveying technology and information, this research focuses on tidal flat digital terrain modeling by neural network. The model structure is 2 hidden layers, 11 neurons in every layer. The terrain calculated is very close to the terrain actual surveyed. RMSE is 0.1564 m. R² is 095817. Residual distribution is normal. The study value is creative to get dynamic tidal flat terrain information fast and efficiently.

Keywords

Accuracy; Artificial neural networks; Digital elevation models; Laser radar; Presses; Simulation; Skeleton; BP Neural Network; digital terrain modeling; the yellow sea radial sand ridges; tidal basin; tidal flat;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Science and Engineering (ICISE), 2010 2nd International Conference on

Conference_Location

Hangzhou, China

Print_ISBN

978-1-4244-7616-9

Type

conf

DOI

10.1109/ICISE.2010.5691689

Filename

5691689